‘Some Hard Decisions Before AI Takes Off’

If there’s one thing John Battelle knows well it’s media — how people digest, create, and transform content across print and virtual platforms. Today we can’t talk about media without addressing data, which Battelle says will underpin nearly every innovative breakthrough — and pause — throughout 2024. That includes AI.

Speaking with Signal360’s Stan Joosten, Battelle walks through his infamous annual predictions for the upcoming year and the big choices brands may want to consider. That includes a suggestion to delve into generative AI, such as smartbots or genies, particularly if their customers are engaging with the technology. But, Battelle warns, be “smart” on how you start.

“It’s nuts to not have a budget for experimentation and learning,” he says. “That’s very different than going all-in on a big project because you want to say in a press release or at a board meeting that you have an AI strategy. But learning and experimenting is crucial.”

You can hear more from this lively conversation between Battelle and Joosten in the video below or read our lightly edited transcript.

TRANSCRIPT:

Stan Joosten
Hello everyone. Welcome to 2024. On behalf of the signal 360. Team, I want to wish you all a very productive, healthy and very happy new year. Welcome, John, Happy New Year to you as well. Good to see you.

John Battelle
Happy New Year Stan. Hi, everybody out there.

I hope you had a very relaxing and somewhat productive break as well, because you’ve been hard at work reflecting on the year past and looking forward. That’s what we want to talk about today. For now, 21 years, which is quite a history, you produce 10 annual predictions on marketing technology and media. You did the same for 2024. But before we dive into 2024 and have a discussion, and we unfortunately can’t cover all 10 of them. But we’ll reflect on 2023. Because it’s was an interesting and somewhat unpredictable year. John, any insights that you want to share with the audience where you reflect back on 2023?

I learned a few things. I should probably stick to technology, and marketing and stay away from anything involving Elon Musk might be the best way of putting it.My batting average went down in 2023. I tried to predict what was happening with Twitter and I had an optimistic take, which is by the end of the year, it would be turned around and that didn’t happen. But when it came to AI, I think what I imagined was going to happen actually did.

I had many predictions on AI, most of which had to do with the way the business would play out. As those out there might recall, a year ago, everyone was freaking out about AI and its impact on search, and how the big players, particularly Google would respond.One of my predictions was that Google would integrate its own AI chatbot, generative AI service, into its main search service, but it would do so very cautiously, which is exactly what played out. And that it would find a business model and start to make some serious money, which also played out as a matter of fact.

When I wrote my predictions review a few weeks ago, the numbers that were being bandied about were that OpenAI was already past a billion dollars in annualized revenue. Just this week, that number has been updated to $1.6 billion in annualized revenue. So my prediction that AI was going to get a business model seems like it is certainly starting to play out.

That’s great. I think that’s the key takeaway from 2023 to AI is here, right? Hopefully, it’s here to stay in some ways. We’ll get back to that to that shortly. We had a lot of drama attached to OpenAI as well, not just Elon Musk. So another thing is to stay away from the drama side of the business. I that might be the recommendation on that one.

You titled your predictions for 2024: “It’s All About the Data.” Can you explain that?

If folks out there have read my predictions, one of the things I do consistently year on year is I don’t think too hard about it. I found that I do better if I don’t overthink it. I just sit down and write. A few hours later, I look up and reread what I’ve written. I found a theme running through almost every one of the things that I’ve noted in my 2024 predictions, and that’s essentially the way that our society is grappling with this new asset, which, broadly speaking, we can tag as data. But what we’re really trying to figure out is, who owns it, who controls it, who governs it? What are the rules? Who gets the value from it? AI has thrown all of this into some very sharp relief. For example, look at the recent lawsuits by The New York Times, last week, and many others in the months prior, and it’s, wait a minute, do you have the right to use my data, my information to train your new models, where you’re going to be banking money in a market that I’m already in business in? Another question is, Hey, shouldn’t I have the right to all the data that I co-create on social media platforms? I should be able to take that data and do other things with it. That is related to one of the predictions I made about how social media networks are going to evolve this year.

Lastly, if we’re going to make a lot of value out of these new tools, we’re going to need as enterprises as large organizations, we’re going to need to be able to tap into govern, manage the data that we have as assets that many large corporations have put billions of dollars into creating over the past couple of decades, as the digital revolution really took hold. Those data assets are very difficult to unlock, because of the way that companies govern them. In many cases, those companies are also in regulatory frameworks, whether it’s financial services, CPG, and there are a lot of regulatory rules around how data might be used. So all of that has to be untangled before you can just throw AI at it and make it all better.

That sets up the conversation nicely across all the predictions. Indeed, that is a nice red line through all of it. Let’s start with AI, which appeared last year. Actually, it was on the horizon for a while, but in public, in the past year, it really became something to behold. Everybody can access ChatGPT. We’ve seen AI win art contests and judge us. We’ve seen the first AI-generated commercials come out. We’ve seen that AI makes for good writers for Sports Illustrated. All kinds of stories around that. Will this story continue the way it has in the past here? What do you see on the horizon? Is this just a steep upward slope here?

I don’t think so. My first prediction is that the AI party is going to take a pause. That doesn’t mean that there won’t be sort of a domination of the headlines by AI-related topics, but rather taken as a whole, we have a lot of digesting to do. This is typical of large narratives that play out across technology over time. Once the general public gets a taste of what might be coming, everyone gets very excited. And once business realizes what might be possible, they start to invest. Then what happens almost always is, “This is going to be harder than I thought or this is going to be very difficult to figure out how to integrate this new technology into a two or three decade old IT stack.” So there’s a lot of integration and digesting that needs to happen. Add to that the complication of the uncertainty of the regulatory frameworks of pending lawsuits of the conservative nature of corporations in terms of not wanting to go too far down a road that might get cut off by new regulations. And  decade now or more where the entire business community has been a little bit burned by getting too far over our skis, when it comes to investing in new technologies, generally, whether that’s web3, and crypto, which was the craze in 2021 to 2022. Or whether it’s, over rotating on investment in social media platforms for marketing and go to market activities, and then realizing that there were downsides to that.

So I think that caution, and a little bit of chin stroking will probably be a big part of 2024, even as we continue to see extraordinary and important work being done to build out the capabilities of this AI ecosystem.

So what do you see on the near horizon? There’s actual tangible applications of AI that might be common place in the next two or three years.

One of the most exciting areas and I’ve been writing about this for some time is what I call the genie. Like literally Aladdin’s Genie that comes out of a lamp and grants you wishes. The idea that consumers might have these genies, which are essentially AI chatbots that are very specifically tuned to do tasks that are otherwise very difficult or time consuming or frustrating or all three things. For example, filing health claims, or trying to book a complicated travel itinerary and ensure that you’re getting the best deal. These are information-driven tasks that require touching a lot of what are called digital surfaces, your data, travel sites, Google Flights, things like that. These are perfect opportunities for entrepreneurs to create new kinds of services, that are tuned to do one thing really, really well. I would expect to see a lot of these over the next 12 to 24 months. And I would expect companies in particular, and marketers specifically to start leveraging these kinds of tools, as customer service capabilities that can be marketed and helped with go to market activities.

That sounds to me like a great opportunity for technology to be truly useful.

That’s right. One of the areas that I look at is the incumbents, the large tech companies who have a lot of wherewithal, because AI is fairly capital intensive, it requires lots of computing power, and requires lots of experts to build it. So I would see, and somewhat proven out the last year, a really nice return for big tech here.

Now you have a bit of a contrary contradictory opinion of what might happen with big tech in the year ahead. I have to say you were a few days ahead of a downgrade by Barclays of Apple. So what’s ahead for big tech? I thought it was all clear sunshine and the next trillion dollars to be added this year for Apple, for example.

Your right. In my predictions, I called it big text midlife crisis. I think we’re already in it for big tech. But even as big tech is starting to age, and it’s getting so big that it’s very hard to continue to grow the way that everyone expects it to grow on Wall Street. In order to grow that way, big tech has to start doing things that essentially aren’t necessarily good for their customers. For example, if you’re an Amazon Prime user, you probably got an email last week, telling you that, “Oh, by the way, we’re going to be adding ads to your Amazon Prime experience.”

Now, of course, Stan, you and I are big fans of advertising at Signal360. But changing the goalposts moving them in this way, is going to make customers angry. If this death by 1000 cuts occurs 10,15, 20 times a month, across four or five large big tech platforms, it starts to make consumers interested in alternatives. I think we’ll see that start to happen this year. I’ve already begun to notice it.

The other thing is is that the regulatory pincher moves by the European Union, various states in the United States, and even the current administration, which is starting to make a lot of noise by regulating AI and big tech. There are a number of actions by the FTC and the DOJ against the big tech companies that are coming into focus this year. All of this plus the fact that big equals slow means for me that these factors will conspire to make sort of a malaise inside the big tech ecosystem. Now, I also predict, to your point about the stock, that there is a significant motivation in terms of the companies themselves to continue to prop up their stock price. So I believe the stocks will continue to do okay, because there are a lot of tools and a lot of levers they can pull, buying back their shares, propping up earnings with things like adding advertising in various places, cutting costs, which of course, all big tech did last year. We will see the quarterly earnings start to lap last year’s bigger numbers after they cut in the second half of 2023. That will probably push the stock price to continue to go up for a while. But I think underneath it, you’ll see consumers start to go, “Wait a minute, I’m not happy with this anymore. I want to look for alternatives.” That’s great for innovation.

How hard is it to innovate in a world where most of our data actually goes through these big tech fee amounts? For example, I have assembled all my Google apps on my phone in a couple of windows and I counted at least 20 that I use in some way. From [Google] Maps to search every day. I have a Google Home set up that knows when I turn off the lights somewhere. And when you talk about the genie, I can imagine that Google will be first to the trough and some others, Apple ,closely behind it. Amazon already has an Alexa. So what chance is there for innovators to play? Or do we require regulatory guidance or intervention to make that innovation possible?

I think this is the year we’ll find the answer to that question, Stan. And it is to me the question, and it’s one of the reasons that I titled the predictions “It’s All About the Data.” Whether or not consumers take control of their own data will, I think depend on an innovation ecosystem, entrepreneurs intrapreneurs, at larger companies doing really interesting things with data. Entrepreneurs aren’t going to en mass all of a sudden decide we want our data back, and we’re gonna go figure out a way to do it by themselves. There has to be a market motivation for it. Sometimes regulatory, you know, relief is the way it’s done. We’ve seen this already passed into law last year in Europe, in the Digital Services Act, and the Digital Markets Act, which I have written about for Signal 360 and I would I would recommend folks get familiar with those pieces of legislation. [They’re] not going to become law anytime soon, in the United States, but they do create a framework for how people might expect to interact with these big tech platforms around the world, even if they’re not inside the European Union. I expect to see some motion on those fronts this year.

On that note to media and marketing, it’s also related to big tech, because the deprecation of cookies is now upon us, right? I think many of the advertisers that I read aren’t totally ready for it yet, and haven’t realized that. How does that relate to big text dominance of advertising?

I feel like it’s been five years now that we we’ve been told that next year is the year the cookie dies. We certainly have had a lot of time in the publishing media and marketing world to get ready for this. But it really is coming. As matter of fact, I saw a story just today that Google has begun to deprecate the cookie, and about 1% of Chrome users worldwide this week. So it’s coming. Are we ready? Yes, I think we are. But there are going to be casualties as the media ecosystem starts to figure out a new way of delivering an identifying value without a cookie. One of those ways has always been that advertisers support publications that have valid connections to important audiences. That’s what a publication really represents, a community of people interested in a particular topic which might be also of interest to the advertiser. That’s why I’m predicting that this year, we’re going to see a resurgence of smaller publications that are not trying to be all things to everyone, but instead, a very strong center of influence for an important market segment. I you think back to what the original Open Internet was like, that’s what the print world represented. We did not build the tools to make it easy to create those kinds of publications when the internet evolved in the 90s in the early 2000s. Then we built an audience extraction machine in the big companies over the last 20 years. I think we’ll start to see the internet get as Anil Dash put it, a little “weird” going forward. By “weird” I really mean is that smaller sites that have passionate communities that marketers just might find worth their time to support.

That’s interesting to hear. Side question, I know that’s near and dear to your heart, because technology has roiled the world of journalism and in quality publications. Is there hope amongst journalist professionals that actually their ranks might be refilled with some quality writing?

I think that one of the core requirements of being a journalist is what I call cynical optimism. In other words, we’re sort of required to be cynical about the world to ask questions to be questioning. However, we also need to be optimistic that people are interested in facts, truth, and the role of information in a healthy society. I do think we are seeing a lot of interesting models emerge to support journalism. They are not all commercial in nature. In other words, they’re not all supported directly by either platform revenues, Facebook, or Google, or even Substack. Nor are they necessarily supported by advertisers. They might be supported by, believe it or not government taxes, and or nonprofits. We’re seeing a lot of innovation in that space. And I’ll be watching it this year.

So will I. By the way, not necessarily all relevant for the business side of things, but I don’t think that you have a fully self-driving Tesla at this time, right?

I wrote a piece probably about seven or eight years ago, saying that the world was not ready for driverless cars, and those who might have been paying attention in 2015, or 2016, there was a lot of hype about how driverless cars were going to take over in a couple of years. That would have been 2018 or 19. I’m quite certain that in a perfect world, driverless cars would be cruising all over the place. The problem is the world’s never going to be perfect. There are questions about driverless cars that as a society, we’re not really ready to grapple with, particularly the regulatory bodies responsible for determining whether or not there will be cars on the road. Those bodies tend to be cities, municipalities, counties, and to ask them to make a moral decision about who’s at fault when a driverless car kills someone, I don’t think they’re going to want to make those decisions. So I don’t think they’re going to let them on the road in any sort of significant way.

In the near term, do you see this as an illustration of the path ahead for AI in general, because a lot of the self driving cars are AI-driven. It’s not like large language models, but it’s still processing lots of data in real time on probabilistic models. So I think we looked a little bit through very optimistic lenses to having a suggested future arrive tomorrow.

Exactly, I think you’re making a really good point. I wish I had tied those two together in my prediction posts. But it is a very good example of the larger issue that I called the digestion issue that we’re going to have with AI. We have a lot of big topics to consider to make decisions about to debate, and, generally speaking these kinds of decisions take time. If you don’t take the time to make them, if you just run ahead and say, “Tech is my true north. Tech is always right. Everything else be damned,” generally speaking, it’s not a good way to do things. We saw that play out from 2012 to 2022. In the world of large technology, advertising driven platforms. We don’t want to necessarily do that again. I think we’re going to be having some hard conversations over the next few years before we see AI truly take off in a way that we can all imagine.

For the last part of it, to’s take a very practical perspective, because a lot of Signal 360’s readers are in the business community who are directly involved in creating the technology on a day-to-day basis. They might be brand managers. What should they take away from your predictions? If you were sitting at P&G as a brand manager, what would you be watching out for and either staying away from or leaning into in the year ahead?

I know that the Jon Moeller, P&G’s CEO is certainly kind of paramount example of this kind of business advice, to stay close to your customer pay attention to what they are doing. If your customers are young, there is no question that your customers are interacting daily with some kind of generative AI. So you need to stay close to that and understand it. Now is the time to get very smart on those consumer behaviors, and to imagine how that impacts the way you go to market, the way that you produce products, the way that you you know, develop products. But also don’t get over your skis. Don’t over invest in something that doesn’t necessarily have a proven return yet, or that has the capacity to change dramatically in the next couple of years. I think that’s going to be a constant, which is a change in how these products are rolled out and how they work. But there is always value in staying connected and close to your core customer set, and to understanding and structuring, and having very good hygiene efforts around your data, the current data that you have, and how you capture and collect new data. By hygiene what I mean is not only that is it usable and structured in a way that is machine readable, but also that it can pass through any kind of compliance or regulatory tests that it needs to. If you aren’t ready to put your data to work upon the opportunity presenting itself, you will miss the opportunity. So focus on structuring and good hygiene with your data.

I think one to add is, answering my own question a little bit here, but experiment to learn, right? The best way to act, to get the questions, is to first understand the questions to ask and then go find and learn some answers. And the best way to do it is by just small experiments, that gets you into what really does.

I think it’s kind of nuts to not have a budget for experimentation and for learning. That’s very different than going all-in on a big project because you want to be able to say in a press release or at a board meeting that you have an AI strategy. But learning and experimenting is crucial.

Before we go, thank you very much for all the insights. But I did notice that you stopped at nine this year nine predictions. Your habit is 10. What if you give the viewers in of this, your 10th? What might be the wildcard for all I care might be who wins the Super Bowl, which will take to Las Vegas, I guess. But what would it be?

I think it will be a very uneventful election year. I didn’t want to put that in the official predictions, because I’ve found that writing about politics is, no one’s interested in that. At least from my point of view. But I think there’s so much at stake that a lot of people are paying attention. And I think we’re gonna have an okay year that way, even though everyone’s freaking out about it.

That will be an interesting starting point for January 5 or so next year. And I invite everybody back to what the predictions might be, and we can look forward. But in the meantime, we have a year ahead and John and I and the Signal360 team are hard at work to lay out a whole content editorial plan that will address this many of the points that John just brought up, and many others. We will appreciate your input as always.

And John, you and I will see each other back often and not in the least because Signal 2024 is planned for July 17. Everybody mark their calendars for Wednesday, July 17. If I’m not mistaken, this afternoon, we have a first planning meeting, to set the theme. More news to come.

I want to thank everybody who joined us. Please feel free to reach out to John, myself or others on the team. And John, I’m looking forward to a very interesting year ahead and let’s take stock of your predictions next year when we get back.

I look forward to it. Thanks, Stan. Appreciate it. Bye everyone.